SUPPLEMENT. 



15 



pressure may then be disregarded, and the effect produed upon the ther- 

 mometers in the compression apparatus must be due to secondary effects 

 of pressure, such as evolution of heat. The various sources of heat were 

 four: 1. Heating of the water by compression. This depends greatly on 

 the original temperature of the water, being nil at the point of maximum 

 density (40° F.) and larger for higher temperatures. One-fourth of the 

 total effect is due to this. 2. Heating of the water due to pumping in 

 through the narrow tube. This accounts for three-twentieths of the 

 effect. 3. Heating of the vulcanite frame by compression. This explains 

 another fifth. 4. Heating due to the effect upon the protecting-bulb. 

 This probably explains the remaining two-fifths of the effect. In this 

 last case, however, there is not only compression but distortion; and of 

 the thermal effects of such a strain no one yet knows anything. These 

 four sources of error cannot be supposed to exist under the conditions in 

 which deep-sea temperatures are taken ; and the only other possible source, 

 that, namely, due to the direct effect of pressure, gives rise to an error 

 which requires a correction of only 0°.04 F. per mile of depth. In 'the 

 course of the description of experiments Professor Tait had occasion to 

 describe the various kinds of pressure gauges which he had found it 

 necessary to devise, the ordinary forms of gauge being altogether useless 

 for scientific work.'" 



